I have just returned from Screwfix/Electrafix to collect an order. I notice in the Electrafix side of the store that there was a poster for electrical stuff showing a BG single module SPD for £24.99. Clearly not a King's ransom.

It would seem to me to be cheap insurance? Certainly much cheaper than most of your electronic filled appliances, the PCB in your boiler and your smoke alarms.

A few years back I lost a modem along with 2 other neighbors after a thunderstorm , We have overhead phone lines. BT master socket with internal surge protection but now have SPD protection on the phone line.

A couple of months back after a storm my TV signal was poor with a lot of pixilation on some weaker stations. Looking outside up at my aerial I installed my self it was pointing uphill rather than previously level as installed. I thought that was the cause of the signal loss also it was an old analogue aerial, 50 ohm comms co-ax not 75 ohm TV stuff and the co-ax no doubt full of water. Not being as brave as I was 25 years ago I thought it was time for a new digital aerial, 75 ohm co-ax installed by a professional. Job done and the aerial fitter measured the signal at the end of the new co-ax before plugging it in to my distribution amplifier saying good signal any more and I would have to install an attenuater. After he left I could see I had an improved signal on the TVs but still some pixilation. I jumped out the amplifier connecting the aerial directly to one TV and no pixilation with more channels than I previously had. Trip to Screwfix to buy and install  a new amplifier thinking the amplifier was spiked. New box fitted TVs now operating better than ever. Opened up the old box to look inside and could see a track between on the PCB between in input socket conection and an adjacent PCB track. I have the photo and have now fitted an SPD to the aerial amplifier.

I already have a Wylex REC 4 isolator with SPD protection inserted in the meter tails. Just need to do something for the recently installed amateur radio 1/2 wave end fed outside antenna.

Happy New Year.

JP

It was a bit careless of you John " I lost a modem along with 2 other neighbors" to lose those two neighbours however I am pleased you got your TV etc sorted out 

Chris Pearson said:

I raised the question of positioning of an SPD at a NAPIT event in the spring. IIRC, somebody had mentioned that protection is provided only for a finite length down the load side. It follows that SPD protection is best placed in (or immediately adjacent to) final distribution boards rather than at the origin.

What is the best way to install a SPD in a consumer unit? It slots into place on the DIN rail but there's more to it than that...

Should it be physically located next to the main switch or is there anything wrong with installing it at the opposite end?

Is it a good or bad idea to install it on the opposite (neutral terminal) side of the main switch to the MCBs / RCBOs?

Some SPD modules connect their live terminal to a busbar but others connect with short lengths of wire. Where should the live wire go if there already is a live wire to RCDs and the end of a busbar for RCBOs inserted into the live output terminal of the main switch? It's a tight squeeze for a second live wire to the SPD!

The positioning of the SPD is based on ensuring that its "leads" are as short as possible, because the inductance of the wires for a rapidly rising voltage like a transient can be very high. So while the voltage across the SPD gets reduced to 1.5kV say, the voltage at the other ends of the leads only gets reduced to 2kV say, due to a 500V inductive drop along the leads.

So what constitutes a part of the SPD's "lead"? Think of the path a normal or earth leak current takes through the L&N or L&E wiring system. Any part of the wiring of the SPD which isn't part of that route constitutes the "leads" of the SPD - it's a conductive path that will have a big voltage drop across it which is lost to the circuit it's supposed to be protecting.

With that in mind, you have to consider which parts of the bus and N/E bars don't form part of the circuit. So if you have a breaker next to the main swich and an SPD at the far end of the bus bar (with connections to the N and E bars arranged similarly) then you're effectively adding the length of the bus bar plus the length of the N or E bar to the SPD lead length. So ideally the SPD wants to be positioned between the main switch and all the breakers. Similarly the SPD N on the N bar should be in position 2, where position 1 is the N from the main switch and positions 3+ are for the circuits. Ditto the E bar.

I think that you need to read section 534. It makes it clear that the leads to the SPD need to be as short as possible. If they are placed in a DB, that means one side or the other of the main switch.

One should also take account of manufacturers' instructions.

In theory yes.

One could argue that any interconnecting wires should not exceed the length of those supplied with the SPD by the manufacturer. It can be assumed that the manufacturer has tested the SPD and it works to spec with the wires supplied with it.

Yeah, but shorter is still better. The spec for the SPD will say something like "with an 8μs rise-time spike, the voltage across the SPD won't exceed 1kV".  With 0.5m leads, that increases to 1.5kV say. What if the rise time of a transient turns out to be 6μs instead? Then the spike may exceed 1.5kV and thus exceed the category II (1.5kV) rating of all those LED bulbs in the house.

Keeping the leads as short as possible, rather than to just whatever the manufacturer supplied, will (in a small way) reduce the probability of a spike which end-user equipment can't deal with.

Take into account that most electronic devices are connected to the CU with significantly longer conductors than those between the main switch and the SPD. Notable exceptions are kWh meters and timeswitches installed inside the CU.

Yes but also note those longer conductors are not in the high current 'crowbar' path trying to shunt the surge, just in the series path transmitting the remaining voltage that was not shunted.
That is the SPD, its tails and all the cable back to the origin of the surge. Ideally only a kV or so is dropped across the SPD, and we take our 'clean' supply for the sensitive loads from as close as we sensibly can to either side of the SPD body. The extra kVs piling up along the wires outside are someone else's problem ;-)

Unless of course the surge is generated at the load end of things, in which case the SPD should be there, not at the CU. The important conductors in terms of the residual voltage drop are the ones through which the bulk of the surge current passes , not any others that happen to connect to high impedance victims. 
Mike

mapj1 said:

Unless of course the surge is generated at the load end of things, in which case the SPD should be there, not at the CU.

Is that (part of) the justification for having SPDs in the final boards? Obviously, it depends upon the nature of the installation, but you wouldn't want a surge sneaking back up a distribution circuit to the main board and then down another distribution circuit whilst the SPD is looking the other way.